+

WO1998026553A1 - Equilibrage de charge et reprise de services de reseau - Google Patents

Equilibrage de charge et reprise de services de reseau Download PDF

Info

Publication number
WO1998026553A1
WO1998026553A1 PCT/US1997/022117 US9722117W WO9826553A1 WO 1998026553 A1 WO1998026553 A1 WO 1998026553A1 US 9722117 W US9722117 W US 9722117W WO 9826553 A1 WO9826553 A1 WO 9826553A1
Authority
WO
WIPO (PCT)
Prior art keywords
service
host
computer
hosts
address
Prior art date
Application number
PCT/US1997/022117
Other languages
English (en)
Inventor
Swee B. Lim
Ashish Singhai
Sanjay R. Radia
Original Assignee
Sun Microsystems, Inc.
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Sun Microsystems, Inc. filed Critical Sun Microsystems, Inc.
Priority to EP97952269A priority Critical patent/EP0978184B1/fr
Priority to DE69738257T priority patent/DE69738257D1/de
Priority to JP52679598A priority patent/JP4087903B2/ja
Publication of WO1998026553A1 publication Critical patent/WO1998026553A1/fr

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
    • G06F9/5061—Partitioning or combining of resources
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
    • G06F11/16—Error detection or correction of the data by redundancy in hardware
    • G06F11/20—Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements
    • G06F11/202—Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements where processing functionality is redundant
    • G06F11/2023—Failover techniques
    • G06F11/2025—Failover techniques using centralised failover control functionality
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
    • G06F9/5083—Techniques for rebalancing the load in a distributed system
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/10—Protocols in which an application is distributed across nodes in the network
    • H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/10—Protocols in which an application is distributed across nodes in the network
    • H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
    • H04L67/1004—Server selection for load balancing
    • H04L67/1008—Server selection for load balancing based on parameters of servers, e.g. available memory or workload
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/10—Protocols in which an application is distributed across nodes in the network
    • H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
    • H04L67/1004—Server selection for load balancing
    • H04L67/1012—Server selection for load balancing based on compliance of requirements or conditions with available server resources
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/10—Protocols in which an application is distributed across nodes in the network
    • H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
    • H04L67/1004—Server selection for load balancing
    • H04L67/1017—Server selection for load balancing based on a round robin mechanism
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/10—Protocols in which an application is distributed across nodes in the network
    • H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
    • H04L67/1034—Reaction to server failures by a load balancer
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
    • G06F11/16—Error detection or correction of the data by redundancy in hardware
    • G06F11/20—Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements
    • G06F11/202—Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements where processing functionality is redundant
    • G06F11/2035—Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements where processing functionality is redundant without idle spare hardware
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
    • G06F11/16—Error detection or correction of the data by redundancy in hardware
    • G06F11/20—Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements
    • G06F11/202—Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements where processing functionality is redundant
    • G06F11/2041—Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements where processing functionality is redundant with more than one idle spare processing component
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2209/00—Indexing scheme relating to G06F9/00
    • G06F2209/50—Indexing scheme relating to G06F9/50
    • G06F2209/505—Clust

Definitions

  • This application relates to systems and methods for management of services within a network of computer systems, and particularly relates to load balancing and host failover within such a network.
  • Horizontal scaleability adds more computer systems as load increases.
  • Each of these computers is typically less powerful than the vertically scaleable solution, but the combined power of multiple such systems frequently exceeds that of the vertical solution for many applications.
  • the horizontal solution also permits the user to maximize the cost benefit of prior investments (i.e., prior purchases of still-compatible computer systems).
  • Horizontal scaleability can therefore be seen to offer a number of-key advantages over vertical scaleability.
  • Reliability is also best served by a horizontally scaleable solution. Reliability is measured by the availability of computer services when needed; since no single computer has ever proved to have 100% up-time, reliability requires more than one computer capable of providing the needed computer services.
  • One-to-one failover designates a host system as primary and a backup system as secondary; in the most classic implementation of this approach, the secondary system is idle - that is, it provides no services - until the host fails. When the host becomes unavailable, the secondary system provides the services normally provided by the host.
  • Symmetric one-to-one failover is a similar technique, wherein each of the "host” and “backup" systems provide distinct but useful sets of services when both are available, and each is capable of providing the services normally provided by the other.
  • each system is both a primary and a secondary, but only the one machine can serve as a backup to the other.
  • the second failover scheme known in the prior art is many-to-one failover.
  • This approach there are many primary systems but only a single secondary, with each of the primaries providing a distinct set of services.
  • the secondary or backup system is capable of performing any of the services provided by any or all of the primaries, but normally sits idle until a primary fails.
  • the present invention overcomes virtually all of the limitations of the prior art by providing a system and method by which computing services may be provided by any of a plurality of systems within a group, such that the computing services provided by that group will continue to be available as long as at least one system is available within the group.
  • the system and method involves assignment of IP addresses to the hosts and services and management of those IP addresses to achieve the functionality described below. To provide many-to-many failover, a plurality of hosts are organized into a
  • Service group wherein each host is capable of providing the service rendered by the service group.
  • Each host may also be available to provide other services, and thus may be members of other service groups.
  • a service is provided by all of the active members of the service group.
  • it is important to maintain communications within the group and to coordinate the cooperation of the various hosts to ensure that the service provided by that group remains available, even if one or more hosts providing the service become unavailable from time to time, such as through failure or load balancing.
  • each host periodically sends a "Control message" to all other hosts within the group.
  • the control message may also be thought of as an "info message” or, for some purposes, a .heartbeat.
  • the control message basically keeps each other host appraised of the status of the sending host as well as data about the perceived status of the other hosts within the group.
  • a Leader host is established and assigns to the various hosts responsibility for particular service addresses.
  • the leader identifies resources, such as service addresses (which may, for example, be IP addresses) which are not being attended and causes such resources to be acquired by designated hosts.
  • the leader also dynamically reassigns resources to hosts which have been newly added or restored to the service group, as well as causing the release of resources handled by failed hosts.
  • Load balancing can also be accomplished with the present invention, either through statistical load-balancing by measuring utilization and using round-robin rotation of entries in a Domain Name Service (DNS) zone, or through DNS zone modification.
  • DNS Domain Name Service
  • Utilization may be determined as a ratio of load (expressed, for example, in instructions per second not idling) divided by current capacity (typically expressed in terms of instructions per second) and for each host is calculated by a service- specific module, since each service supported by a host may have different load and capacity.
  • load shedding can be initiated if a hosts utilization exceeds that high water mark by shifting new connections to other less-loaded hosts.
  • a hosts utilization drops below a preset minimum utilization (or low water, mark), that host can be added back to the E)NS zone.
  • DNS zone modification can be used for load balancing by establishing a configuration parameter that specifies the minimum number of available service addresses that must be present in a DNS zone. By such an approach, highly loaded hosts that exceed the high water mark may continue to be identified as available.
  • FIG. 1A shows in block diagram form the operation of a RAIS daemon within a host handling various object types in accordance with the present invention.
  • Figure 1 B shows the linking of various service monitors to their associated test objects and load objects.
  • Figure 2A shows in block diagram form a network comprising a plurality of service groups.
  • Figure 2B shows the data structure of a Control Info Message.
  • Figure 2C shows the data structure of the Host Information Entry.
  • Figure 2D shows the data structure of the Service Address Database.
  • FIG. 3 shows in flow diagram form the "transmit" thread of the service monitor.
  • Figure 4A shows in flow diagram form the "receive, thread of the service monitor.
  • Figure 4B shows in flow diagram form the "acquire message portion of the receive thread.
  • Figure 4C shows in flow diagram form the Receive message portion of the receive thread.
  • Figure 4D shows in flow diagram form the "update tables" portion of the receive thread.
  • Figure 5A shows in flow diagram form the "scripts" thread of the service monitor.
  • Figure 6B shows the data structure of the control service address table.
  • Figure 6C shows the data structure of a table showing utilization and weighted utilization.
  • the present invention provides a system and method by which a computing service may be provided by any of a plurality of systems within a group, such that the computing service provided by that group will continue to be available as long as at least one system is available within the group.
  • the invention is particularly well-suited to operation on a Redundant Array of Independent Servers, or RAIS.
  • IP addresses of a plurality of computer systems each of which may be thought of as a service provider for a computing service available from the network.
  • Each computer system, or host is assigned a unique IP address in the conventional manner, but the IP addresses are then grouped according to providers of a desired service, such that each member of the group is available to provide that service.
  • This collection of hosts may be referred to a "service group, and the members of a service group cooperate to provide a particular service.
  • a host may belong to multiple service groups, but for purposes of clarity and simplicity the following discussion will be limited to describing the interaction among hosts in a single service group.
  • a "service group” address is established, which is simply an IP address for that group. All members of the service group communicate with each other using the service group address - that is, each member transmits messages to all other group members by addressing the message to the service group address, and each member receives messages from all other service group members by receiving messages destined to the service group address.
  • These messages which may be thought of as "control messages,” communicate the well-being of the host providing the message, the status of each service address, actions to be performed by the members of the service group, and any other function suited to the messaging capability.
  • a Control address refers to the IP address of a control interface. Each host will preferably have one or more control addresses, and typically must have at least one control interface. The first control address is typically used to identify uniquely the associated host, and may also be referred to as the "host id” or “id_address” hereinafter. Control addresses identify which interface the host should use to transmit and receive group messages.
  • a service address is an IP address that hosts in the service group provide service through. Management of such service addresses is a key portion of the present invention.
  • each host within a service group periodically sends control (or info) messages to the service group via the service group address.
  • info message includes, among other things, a set of the hosts control addresses, its preferred service address and a set of service addresses it is currently serving. These messages enable members of the service group to determine the status of each host in the service group.
  • Another function of the control messages is to permit the hosts within a service group to elect one of the hosts to fill the logical role of "leader" of the service group. Because the role of leader is logical, the role may move from one host to another at various times, including when hosts are added or removed. In an exemplary embodiment, the host with the lowest host id is denominated the leader, although other methods are also acceptable.
  • the leader of a service group manages service addresses within the group. Among other tasks, the leader tries to establish various conditions at equilibrium, including that all service addresses are being served, that each service address s assigned to only one host, and that each host will serve its preferred service address. The leader also initiates corrective action if the conditions are violated, typically by causing a host either to serve or stop serving a service address; this may be thought of as "acquiring" or "releasing" By such acquiring and releasing Of service addresses, the hosts included within a service group ensure that all addresses are being served. In addition, the present invention includes the ability to modify domain name service bindings by modifying zone files and notifying the DNS server of zone file changes, as described in the concurrently filed U.S.
  • a zone is a group of IP addresses with a single common parent, although a zone may not include all of the child addresses or any lower level nodes.
  • the leader is also solely responsible for assigning service addresses; thus, while a host added to the system presents with a preferred service address, it does not own that service address until the leader assigns it.
  • the leader listens to control messages and quickly releases the preferred service address introduced by the joining host if currently being served by another host by another host in the service group. If the preferred service address is not being served by anyone, the leader simply assigns it to the new host.
  • the case where a different hoot has already been assigned the preferred service address of the new host usually arises where a host has previously failed, and is now rejoining the group. In such a situation, the leader will typically request the host currently owning the preferred service address to release it, and will then wait for the release request to complete before assigning the preferred service address to the joining host.
  • the leader When a host fails, the leader will detect that the service addresses that were originally served by the failed host are not being actively served, because the control messages will show increasingly long periods since the last time the service addresses were served. The leader then reassigns these unserved addresses to other available hosts within the group.
  • the reassignments may be made randomly, or may be made base on a load-balancing scheme, or through any other suitable method.
  • the reassignment of service addresses permits failed hosts to be removed from the service group.
  • a service address that is not assigned to its preferred host is referred to as an "orphaned" service address.
  • Associated with each orphaned service address is a logical countdown timer maintained by the leader. This countdown timer is created and started when the address becomes orphaned; when H expires, the service address is invalidated. When the service address goes invalid, the leader notices that a host is serving an unknown service address and requests that host to release it.
  • the countdown timer can be set for very high values, perhaps in the range of years.
  • the integrated name service mentioned above refers to the ability of the present invention to modify DNS bindings by modifying zone files and notifying the DNS server of the zone file changes. See the applications described in the Related Applications portion, above; - in particular, see U.S. Pat. ppn. S.N. 08/673,951 , entitled "A Service For a Redundant Array of Internet Servers" of Swee Boon Lim, filed 7/1/96.
  • each service group has its own DNS zone. This requirement isolates the DNS changes originating from a service group to a single DNS zone file.
  • the DNS zone is set up such that only preferred service addresses of available members are advertised. Since orphaned service addresses cannot be obtained, orphaned service addresses will fade away as described above. Load balancing is another important aspect of the present invention.
  • the info message includes two values: one is the current load; the other is current capacity.
  • Current capacity is typically expressed in instructions per second; current load is typically expressed in number of instructions per second not idle, measured over a suitable duration.
  • capacity could be the maximum bandwidth of the service interfaces, while load may be the average bandwidth consumed during the last measurement interval.
  • Hot sparing involves hosts designated as service providers only if another host fails; in essence, such devices are surrogates for failed hosts and their preferred service addresses.
  • a hot spare host does not have a preferred service address, but instead sends info messages advising that it is available. This allows the leader to assign service addresses to the hot spare in the event it is needed. In particular, the leader will typically assign orphaned service addresses to hot spares only. Assignment of service addresses may either be random or in accordance with a load-balancing scheme.
  • Quiescing is defined as preventing a host in a service group from acquiring service addresses, while at the same time permitting that host to release addresses; in an exemplary embodiment, setting a host as quiesced requires intervention by the system operator.
  • quiesced hosts are not included in the service groups DNS zone, and thus their address does not show up in the list of available hosts; unlike failed hosts a quiesced host can continue to serve its current service addresses.
  • Failure detection also available to be performed by the present invention, involves monitoring of the "info" or control messages sent by each host.
  • the info messages basically serve as a "heartbeat" signal.
  • a total host failure is detected if no heartbeat signal occurs within a timeout interval.
  • the frequency of transmission of the info message/heartbeat and the timeout interval may be adjusted by the operator to optimize system operation.
  • a host may experience service failures, which occur when a host can no longer provide a service due to partial failure. These failures may be transient or persistent, and may involve software or hardware failures such as resource exhaustion or corrupted data, among others.
  • hosts with service failures may continue to participate in a service group, although such hosts will voluntarily release their service addresses.
  • service monitor SM1 and SM2 and the associated test objects T01-T02 and load objects L01 and L02 are illustrated in the context of the RAIS daemon 20.
  • service monitor SM1 is linked to test objects T01 and T02 and load object LOT
  • service monitor SM2 is linked to test object T03 and load object L02.
  • a service monitor may be linked to any number of test objects, and each test object and load object may be linked to more than one service monitor.
  • each provider of a service 30 is identified by a unique IP address.
  • each network service 30 may be provided by a plurality of hosts 10 of the type shown in Figures 1A-1 B. This plurality of hosts is configured in what is referred to herein as a service group 100, and multiple service groups 100 may be interconnected within a network 110 to provide all necessary computer services.
  • a single host 10 may belong to multiple service groups 100; however, for purposes of clarity, the present description will describe the interaction among hosts in a single service group.
  • a new IP address referred to as a service group address 105 is established for each service group, and is typically an IP multicast address and a port number. All hosts 10, or members, within a service group 100 communicate with one another using the service group address 105 that is each member transmits messages to all other group members by addressing the message to the service group address, and each member receives messages from all other service group members by receiving messages destined to the service group address. These messages, which may be thought of as control messages, are typically broadcast on a periodic basis by each host using a protocol such as UDP and communicate a variety of data. The data structure of the control/info message is shown in Figure 2B.
  • Typical message contents include: (1) the well-being of the host providing the message; (2) capacity and load; (3) a list of control addresses; (4) the preferred service address; (5) a list of addresses currently being served, including (a) the service address, and (b) the state of that service address, which can be either acquiring, releasing, on-line, or not served; and (6) a list of addresses that should be served including (a) the service address and (b) the last time that address was served by the preferred host.
  • the foregoing data about the transmitting host which is included in the control message is typically maintained in a data table referred to as a Host Information Entry, and is transmitted only if authoritatively correct.
  • An exemplary Host Information Entry structure is shown in Figure 2C.
  • control message further includes whatever information that host has received about the other hosts included in the service group, and when information was last received from that other host (i.e., aging.)
  • This data is stored is a database, referred to as the Host Information Database.
  • the data structure of Host Information Database is simply a series of Host Information Entries, organized by id_address.
  • the Control Message includes a Service Address Database, which is arranged as a set of service addresses and the time since last update for each; See Figure 2D for an exemplary structure. As will be appreciated hereinafter, the time is reset with every new entry; and such resets are used to ensure that the host responsible for that service address has not died. A failed, or dead, host stops sending messages, which alerts the present invention to reassign that hosts tasks.
  • the transmission of the control messages is handled by the Service
  • the Service Monitor associated with each service, such as SM1.
  • the Service Monitor is multithreaded, for which the threads are created for: Transmit, Receive, Script, Control and DNS.
  • the Transmit thread transmits the control message, while the Receive thread receives control messages from other hosts within the service group.
  • the Script thread handles all events that may take too long for handling by the other threads, while the control thread determines which of the hosts within the service group will be "leader" and will make assignment decisions.
  • the DNS thread simply updates the DNS. Each of these threads will be discussed in greater detail hereinafter. Referring next to Figure 3, the Transmit thread may be better understood.
  • each host begins one Transmit thread for each service.
  • the Transmit thread starts at step 300 in the "Sleep" state.
  • the process advances at step 305 by invoking the test objects TO .
  • On associated with that service followed at step 310 by checking whether the test is passed. If the tests fail, or a NO results at step 310, the process advances to step 320 and invokes a script to schedule the release of all current service addresses (so that the service addresses can be reassigned to another, non-failed host), after which a control message is created and sent at step 330.
  • the process then loops back to sleep at step 300.
  • step 350 load and capacity information (discussed above) is gathered from each of the load objects L01...L0n.
  • the service address and timer data is then written if an Im alive time-out timer requires it.
  • a control or info message is then created and sent at step 370 to communicate the state of the host to the remaining members of the service group, and at step 380 ARP packets are sent.
  • the process then loops back to sleep at step 300. Thereafter, the daemon will restart the process after an appropriate time.
  • the Receive thread is shown at a top level in Figure 4A, with Figures 4B-4D showing the various branches that are included as part of the Receive thread.
  • the Receive thread initiates at step 400, and advances to step 402 where a message is received from another host within the service group. A check is then made at step 404 to determine whether the message is a duplicate; if so, the process loops back to step 402 to receive another message. If the message is not a duplicate, the process advances to step 406 and checks to see whether the message is an acquire message. If so, the process branches to Figure 4B, discussed below. If not, the process advances to step 408 and checks to see whether the message is a release message.
  • step 410 a check is made to determine whether the message is an info, or control, message. If so, the process branches to Figure 4D, also discussed below. If not, the message is discarded and the process loops back to step 402.
  • the Acquire process basically involves having the leader send a message identifying the source and destination id_addresses, together with which service address is to be acquired.
  • the process starts at step 412 by checking whether the received acquire, message is for that host. If not, the process is done and returns at step 414 by looping to step 402. If the message is for that host, the process advances to check whether the message is from that hosts leader by comparing addresses in the Host Information Database. Again, if not, the process again returns at step 414. If the check at step 416 yields a YES, the process advances at step 418 and determines whether the receiving host is OK.
  • step 420 a check is made to determine whether the host is quiesced (which can be ascertained from the Host Information Entry, although operator intervention is typically required to set the host as quiesced.) If the answer at step 420 is yes, no service address can be acquired, so the process again returns at step 414. However, if the host is not quiesced, a check is then made at step 422 to determine whether the host currently is servicing the service address identified in the acquire message. This can be determined by examining the hosts Host Info Entry. If so, there is no need to reacquire the "new" service address, and the process returns at 414.
  • the process advances at step 424 by updating the Host Information database with an entry for the new service address, and an acquire action is scheduled at step 426.
  • the "acquire" action is described in greater detail in connection with Figure 5B. The process then returns at step 414.
  • the Release portion of the Receive thread can be better appreciated. From the following, it will be apparent to those skilled in the art that the Release portion is substantially similar to the Acquire portion.
  • the process starts at step 432 by checking whether the message is for the receiving host. If not, the process returns at step 434. If so, the process advances to step 436 and checks whether the message is from the leader of that service group by comparing Host IDs in the Host Information database; if not, the process again returns at step 434. If the message is from the leader, then a check is made at step 440 (by looking at the relevant flag in the Host Info Entry) to determine if the identified service address to be released is on-line.
  • the "Release message comprises source and destination id_addresses. If the service address to be released is not being serviced by the receiving host, the process returns at step 434. However, if the service address is being serviced, as determined by the listing in the Host Info Entry, the service address is first marked as "releasing" in the Service Address Information, and the Host Information database is updated at step 442 by deleting that service address from the list of service addresses serviced by that host, and the process advances to step 444 to schedule a release action. A release action is described in connection with Figure 5C. The process then returns at step 434.
  • the response to an info, or control message can be understood.
  • the response to an info message is simply to update the appropriate tables maintained by the receiving host, particularly the Host Information database referred to previously, which effectively updates the status of each service address.
  • the process begins at step 450 by updating the Service Address database with the information received from the other host, and then advances to step 452 by updating table, or list, of addresses that should be served by the service group.
  • the list specifies, for each service address, the last time that address was served by the preferred host, and resets a time-out timer that basically serves to confirm that the responsible host is alive and well. Once these tables are updated, the process returns at step 454.
  • FIG. 5A shows the overall thread
  • Figures 5B and SC show key branches of the thread.
  • Scripts are generally more complicated processes, and because of this may be run after some delay rather than immediately.
  • the process begins at 500 and advances to step 502 by removing the request for action from the queue.
  • a check is then made at step 504 to determine whether the requested action is an "Acquire" action. If so, the process branches to Figure 5B. If not, the process advances to step 506 to determine whether the requested action is a "Release" action.
  • step 508 a check is made to determine whether the action requested is a test failed action. If the check results in a YES, a "Test Failed" SNMP trap is sent at step 510. If the check results in a NO, the process advances to step 512 where a check is made to determine whether the requested action is a "Cannot Assign" action. If the check results in a YES, the process branches to step 514 and a "Cannot Assign" SNMP trap is sent. If the answer is a no, the process completes at step 516. Turning next to Figure 5B, the "Acquire" branch from Figure 5A may be better understood.
  • the acquire process assigns a service address to a host.
  • the branch begins at step 530 by configuring the hardware service interface for the software service address, and then advances to step 532 where the Acquire script is run.
  • the process advances to step 534 and a check for errors is made. If no errors occurred, the process advances to step 536 and updates, for that service address, the Host Info Entry to On-line.
  • the process then advances to step 538 and sends an SNMP Acquire trap, following which the process completes at step 540. If an error occurred at step 534, such that a YES was returned, the process advances to step 542 by configuring the service interface hardware to remove the service address.
  • step 544 the process then continues at step 544 by updating the Host Info Entry for that service address to "not serving" and then, at step 546, sends the "Acquire fail" SNMP trap.
  • the process then completes.
  • the Release script branch of the Scripts thread can be better appreciated. Similar to the Acquire branch shown in Figure 5B, a release action causes a host to release one of the service addresses it has been handling. The branch begins at step 560, and runs the Release script. The service interface (hardware! is then configured at step 562 for the service address (software) to remove the service address being released.
  • a check for errors is made at step 564, and if an error occurs a Release Fail trap is sent at step 566, after which the process completes at step 568. Whether or not an error occurs, a "Release" trap is sent at step 570 the Host Info Entry is also updated to remove the affected Service Address and the process completes.
  • the Control thread of the service monitor can be better understood. The purpose of the Control thread is to establish which host within the service- group will be the leader, and to permit the leader to manage service addresses dynamically in accordance with the state of the system.
  • the process starts at step 600 in the Sleep mode. When awakened, the processes advances at step 602 to determine who the leader is currently A check is made at step 604 to determine whether the current host is the leader If not, the process loops back to step 600.
  • step 606 removes old, timed-out entries and silent host information by updating the database to the current state.
  • a silent host is a host that has failed to send a control info message within the time-out period, such that the leader assumes the host to have died. This permits additional decisions, such as reassignment of the service addresses formerly served by a failed host to be made reliably by the leader.
  • a control database is built at step 608, discussed in greater detail below. After the control database has been built, each of the service addresses is processed at step 610 to update its status to current information, also described in greater detail below. The process then loops back to the Sleep mode at step 600.
  • the control database basically comprises three tables.
  • the first is the control service address table, shown in Figure 6B, which sets forth four items: (1) the service address within the service group, (2) the states of each service address (i.e., acquiring, releasing, on-line, or unserved), (3) the preferred service address (or preferred id_address) of each host, and (4) a set of id_addresses which claim to service that service address.
  • the second table is the control Spare_Host_List and the third is the Non_Spare_Eligible_List; these are simply a single-column list of IP addresses. For hot spares, the preferred service address is operator-configured to zero, or not assigned.
  • the table is built through a pair of nested loops, beginning with an examination of the Host Information database.
  • a pseduo-code representation for the outer nested loop of an exemplary embodiment is:
  • id_address means the Host ID.
  • id_address means the Host ID.
  • a- check must be made by the leader for duplicate hosts serving the same service address, and one of the duplicates removed.
  • all control decisions are made by the leader.
  • the foregoing process can be seen to assign the service address to the preferred service address if that address is available; otherwise, the leader assigns the service address to an available hot spare. If neither of these works, the leader can simply pick one. The leader then tells the remaining service addresses to release the address. An exception to this process exists for the artificially injected address discussed above; in this instance, the artificially inserted address is not told to release.
  • the final thread of the Service Monitor is the DNS thread.
  • the DNS thread basically determines load versus capacity for load balancing, and is calculated from a comparison the utilization of the various service addresses both currently and historically.
  • a first table or database M shown in Figure 6C, stores the historical utilization and weighted utilization for each service address, while a second table or database BB stores current utilization and weighted utilization for the various service addresses.
  • databases M and BB have the same structure, with the data from database BB simply being transferred to database M as a new sample is entered in database BB.
  • the pseudocode representation of the DNS thread can then be represented as follows:
  • index starts at minimum of service address and ends at number of entries in BB) if (entry's weightedjjtil > high water mark) then remove entry endif endfor endif if not same list addresses as M then update DNS zone endif swap BB and M [i.e., current becomes last and last will become next current
  • the scaling factor can be varied according to optimize performance according to a number of criteria.
  • the scaling factor is typically in the range of 100 to 1000, so that the value of "Util" is an integer.
  • the scaling factor, once selected, will typically remain fixed although some paradigms exist in which variation of the scaling factor may be desirable.
  • the value of Delta is the difference between the high water mark and the low water mark, and is preset.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Software Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Computer Hardware Design (AREA)
  • Quality & Reliability (AREA)
  • Computer And Data Communications (AREA)
  • Maintenance And Management Of Digital Transmission (AREA)
  • Multi Processors (AREA)

Abstract

Système et procédé pour la reprise et l'équilibrage de charge multivoques, permettant d'établir plusieurs groupes de services pour la fourniture de services de calcul voulus. Chaque groupe de services comprend plusieurs hôtes, chaque hôte du groupe étant disponible pour assurer les services de calcul pour lesquels le groupe dans sa totalité est responsable. Chaque hôte peut appartenir à plusieurs groupes de services. Chaque hôte opérationnel au sein d'un groupe de services transmet des messages périodiques à un autre hôte quelconque au sein du groupe de services, informant de l'état de l'hôte émetteur. Un hôte de tête évalue les messages périodiques et, lorsque c'est approprié, réattribue de manière dynamique la responsabilité de services de calcul particuliers à un hôte au sein du groupe. La réattribution peut être due soit à une reprise soit à un équilibrage de charge.
PCT/US1997/022117 1996-12-09 1997-12-06 Equilibrage de charge et reprise de services de reseau WO1998026553A1 (fr)

Priority Applications (3)

Application Number Priority Date Filing Date Title
EP97952269A EP0978184B1 (fr) 1996-12-09 1997-12-06 Equilibrage de charge et reprise de services de reseau
DE69738257T DE69738257D1 (de) 1996-12-09 1997-12-06 Lastausgleich und ausfallübergabe von netzwerkdiensten
JP52679598A JP4087903B2 (ja) 1996-12-09 1997-12-06 ネットワーク・サービスの負荷平衡化及びフェールオーバ

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US08/763,289 1996-12-09
US08/763,289 US5938732A (en) 1996-12-09 1996-12-09 Load balancing and failover of network services

Publications (1)

Publication Number Publication Date
WO1998026553A1 true WO1998026553A1 (fr) 1998-06-18

Family

ID=25067406

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US1997/022117 WO1998026553A1 (fr) 1996-12-09 1997-12-06 Equilibrage de charge et reprise de services de reseau

Country Status (5)

Country Link
US (1) US5938732A (fr)
EP (1) EP0978184B1 (fr)
JP (1) JP4087903B2 (fr)
DE (1) DE69738257D1 (fr)
WO (1) WO1998026553A1 (fr)

Cited By (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0990986A2 (fr) * 1998-09-30 2000-04-05 Ncr International Inc. Rétablissement de défaillance de systèmes d'ordinateurs partagés contenant un schéma de base de données
EP0992909A2 (fr) * 1998-09-30 2000-04-12 Ncr International Inc. Provision de service de base de données et performance d'interrogation échelonnable par l'usage des copies redondantes
WO2000031942A3 (fr) * 1998-11-20 2000-10-05 Network Alchemy Inc Procede et appareil pour un processus tcp/ip d'equilibrage et de reprise dans un systeme de mise en grappe de reseau de protocole internet (ip)
US6341356B1 (en) 1999-03-25 2002-01-22 International Business Machines Corporation System for I/O path load balancing and failure which can be ported to a plurality of operating environments
WO2001080003A3 (fr) * 2000-04-17 2002-06-06 Circadence Corp Systeme et procede de decalage de fonctionnalites entre des serveurs web multiples
DE10134492B4 (de) * 2000-07-28 2006-02-09 International Business Machines Corp. Ausfallübernahme des Dateimanagementsystems in einem Rechnercluster
EP1061710A3 (fr) * 1999-06-17 2006-08-30 International Business Machines Corporation Système et méthode de répartition de charge intégrée et de gestion de ressources dans un environnement Internet
US7890450B2 (en) 2007-03-16 2011-02-15 Fujitsu Limited Policy creating apparatus, policy creating method, and computer product
USRE45009E1 (en) 2000-04-17 2014-07-08 Circadence Corporation Dynamic network link acceleration
US8898340B2 (en) 2000-04-17 2014-11-25 Circadence Corporation Dynamic network link acceleration for network including wireless communication devices
US8923757B2 (en) 2012-08-23 2014-12-30 Blackberry Limited Method, system and apparatus for service loss handling
US8996705B2 (en) 2000-04-17 2015-03-31 Circadence Corporation Optimization of enhanced network links
US9088588B2 (en) 2012-11-28 2015-07-21 At&T Intellectual Property I, Lp Method and apparatus for registering communication devices in a communication system
US9148293B2 (en) 2000-04-17 2015-09-29 Circadence Corporation Automated network infrastructure test and diagnostic system and method therefor
US10033840B2 (en) 2000-04-17 2018-07-24 Circadence Corporation System and devices facilitating dynamic network link acceleration

Families Citing this family (147)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH09245007A (ja) * 1996-03-11 1997-09-19 Toshiba Corp 情報処理装置及び情報処理方法
US6101542A (en) * 1996-07-19 2000-08-08 Hitachi, Ltd. Service management method and connection oriented network system using such management method
US7273601B2 (en) * 2000-07-18 2007-09-25 The University Of Western Ontario Preparation of radiolabelled haloaromatics via polymer-bound intermediates
FR2773428B1 (fr) * 1998-01-06 2000-02-04 Bull Sa Procede de communication dans un ensemble de systemes distribues via un reseau du type internet
GB2334353B (en) * 1998-02-12 2002-08-07 Ibm An apparatus,method and computer program product for client/server computing with the ability to select which servers are capable of creating transaction stat
US6112311A (en) * 1998-02-20 2000-08-29 International Business Machines Corporation Bridge failover system
US7401114B1 (en) * 1998-04-20 2008-07-15 Sun Microsystems, Inc. Method and apparatus for making a computational service highly available
US6115752A (en) * 1998-05-21 2000-09-05 Sun Microsystems, Inc. System and method for server selection for mirrored sites
US6311217B1 (en) * 1998-06-04 2001-10-30 Compaq Computer Corporation Method and apparatus for improved cluster administration
US6138159A (en) * 1998-06-11 2000-10-24 Phaal; Peter Load direction mechanism
US6078960A (en) * 1998-07-03 2000-06-20 Acceleration Software International Corporation Client-side load-balancing in client server network
US6327622B1 (en) 1998-09-03 2001-12-04 Sun Microsystems, Inc. Load balancing in a network environment
US6324580B1 (en) 1998-09-03 2001-11-27 Sun Microsystems, Inc. Load balancing for replicated services
US6092178A (en) * 1998-09-03 2000-07-18 Sun Microsystems, Inc. System for responding to a resource request
US7206849B1 (en) * 1998-10-05 2007-04-17 Symbol Technologies, Inc. Communication in a wireless communications network when a mobile computer terminal may be unreachable
US6765919B1 (en) * 1998-10-23 2004-07-20 Brocade Communications Systems, Inc. Method and system for creating and implementing zones within a fibre channel system
US6834302B1 (en) * 1998-12-31 2004-12-21 Nortel Networks Limited Dynamic topology notification extensions for the domain name system
US6507863B2 (en) * 1999-01-27 2003-01-14 International Business Machines Corporation Dynamic multicast routing facility for a distributed computing environment
US6412001B1 (en) 1999-01-29 2002-06-25 Sun Microsystems, Inc. Method to monitor and control server applications using low cost covert channels
US6745241B1 (en) * 1999-03-31 2004-06-01 International Business Machines Corporation Method and system for dynamic addition and removal of multiple network names on a single server
US6516350B1 (en) * 1999-06-17 2003-02-04 International Business Machines Corporation Self-regulated resource management of distributed computer resources
US6446218B1 (en) * 1999-06-30 2002-09-03 B-Hub, Inc. Techniques for maintaining fault tolerance for software programs in a clustered computer system
US6374297B1 (en) * 1999-08-16 2002-04-16 International Business Machines Corporation Method and apparatus for load balancing of web cluster farms
JP3636947B2 (ja) * 1999-08-31 2005-04-06 株式会社日立製作所 情報サービスシステム、サービス利用クライアント及びサービス規制方法
US6667980B1 (en) * 1999-10-21 2003-12-23 Sun Microsystems, Inc. Method and apparatus for providing scalable services using a packet distribution table
US6735206B1 (en) 2000-01-10 2004-05-11 Sun Microsystems, Inc. Method and apparatus for performing a fast service lookup in cluster networking
US6587866B1 (en) 2000-01-10 2003-07-01 Sun Microsystems, Inc. Method for distributing packets to server nodes using network client affinity and packet distribution table
US6862613B1 (en) 2000-01-10 2005-03-01 Sun Microsystems, Inc. Method and apparatus for managing operations of clustered computer systems
US6735205B1 (en) 2000-01-10 2004-05-11 Sun Microsystems, Inc. Method and apparatus for fast packet forwarding in cluster networking
US6938256B2 (en) 2000-01-18 2005-08-30 Galactic Computing Corporation System for balance distribution of requests across multiple servers using dynamic metrics
JP3833117B2 (ja) * 2000-01-31 2006-10-11 富士通株式会社 サーバ決定方法及び装置
US7054948B2 (en) * 2000-03-07 2006-05-30 Opcoast Llc Collaborative host masquerading system
US6732189B1 (en) * 2000-03-20 2004-05-04 International Business Machines Corporation Method and apparatus for fault tolerant tunneling of multicast datagrams
US6898727B1 (en) * 2000-03-22 2005-05-24 Emc Corporation Method and apparatus for providing host resources for an electronic commerce site
US20010047387A1 (en) * 2000-03-27 2001-11-29 Exoplex, Inc. Systems and methods for providing distributed cross-enterprise portals
US7266490B2 (en) 2000-12-28 2007-09-04 Robert Marc Zeidman Apparatus and method for connecting hardware to a circuit simulation
US8160863B2 (en) * 2000-03-28 2012-04-17 Ionipas Transfer Company, Llc System and method for connecting a logic circuit simulation to a network
US6993587B1 (en) * 2000-04-07 2006-01-31 Network Appliance Inc. Method and apparatus for election of group leaders in a distributed network
US6701449B1 (en) 2000-04-20 2004-03-02 Ciprico, Inc. Method and apparatus for monitoring and analyzing network appliance status information
US6922724B1 (en) 2000-05-08 2005-07-26 Citrix Systems, Inc. Method and apparatus for managing server load
EP1297412B1 (fr) * 2000-05-08 2008-04-16 Citrix Systems, Inc. Systeme et procede d'envoi de donnees operationnelles dans un reseau de serveurs
US6785726B1 (en) 2000-05-08 2004-08-31 Citrix Systems, Inc. Method and apparatus for delivering local and remote server events in a similar fashion
US6785713B1 (en) * 2000-05-08 2004-08-31 Citrix Systems, Inc. Method and apparatus for communicating among a network of servers utilizing a transport mechanism
US6789112B1 (en) 2000-05-08 2004-09-07 Citrix Systems, Inc. Method and apparatus for administering a server having a subsystem in communication with an event channel
US7280529B1 (en) * 2000-05-20 2007-10-09 Ciena Corporation Providing network management access through user profiles
US7844513B2 (en) 2000-07-17 2010-11-30 Galactic Computing Corporation Bvi/Bc Method and system for operating a commissioned e-commerce service prover
KR100840960B1 (ko) 2000-07-17 2008-06-24 갈랙틱 컴퓨팅 코포레이션 비브이아이/아이비씨 동적 호스트 서비스 관리를 제공하기 위한 방법 및 시스템
US6816905B1 (en) 2000-11-10 2004-11-09 Galactic Computing Corporation Bvi/Bc Method and system for providing dynamic hosted service management across disparate accounts/sites
US8538843B2 (en) 2000-07-17 2013-09-17 Galactic Computing Corporation Bvi/Bc Method and system for operating an E-commerce service provider
AU2000266395A1 (en) * 2000-08-11 2002-02-25 George Mount Server-side session management
US7194549B1 (en) * 2000-09-06 2007-03-20 Vulcan Patents Llc Multicast system using client forwarding
DE20017445U1 (de) * 2000-10-11 2001-03-08 Wall, Georg, 81375 München Verbindungselement für Spundbohlen
US6813635B1 (en) * 2000-10-13 2004-11-02 Hewlett-Packard Development Company, L.P. System and method for distributing load among redundant independent stateful world wide web server sites
US7606898B1 (en) 2000-10-24 2009-10-20 Microsoft Corporation System and method for distributed management of shared computers
US7113900B1 (en) 2000-10-24 2006-09-26 Microsoft Corporation System and method for logical modeling of distributed computer systems
US6938122B2 (en) * 2001-01-23 2005-08-30 Emc Corporation Remote mirroring in a switched environment
US8176137B2 (en) * 2001-01-31 2012-05-08 Accenture Global Services Limited Remotely managing a data processing system via a communications network
US7389341B2 (en) * 2001-01-31 2008-06-17 Accenture Llp Remotely monitoring a data processing system via a communications network
JP2002244870A (ja) * 2001-02-19 2002-08-30 Hitachi Ltd システム管理支援方法及び装置
US20020143946A1 (en) * 2001-03-28 2002-10-03 Daniel Crosson Software based internet protocol address selection method and system
US7366194B2 (en) 2001-04-18 2008-04-29 Brocade Communications Systems, Inc. Fibre channel zoning by logical unit number in hardware
US7113993B1 (en) * 2001-06-20 2006-09-26 Microstrategy, Inc. Technique for handling server session requests in a system having a plurality of servers
US6922791B2 (en) * 2001-08-09 2005-07-26 Dell Products L.P. Failover system and method for cluster environment
US7305674B2 (en) * 2001-08-31 2007-12-04 International Business Machines Corporation Method and apparatus to manage multi-computer supply
US7039916B2 (en) * 2001-09-24 2006-05-02 Intel Corporation Data delivery system for adjusting assignment of connection requests to nodes based upon the tracked duration
US6938031B1 (en) 2001-10-19 2005-08-30 Data Return Llc System and method for accessing information in a replicated database
US7000016B1 (en) 2001-10-19 2006-02-14 Data Return Llc System and method for multi-site clustering in a network
US6811085B2 (en) * 2001-10-26 2004-11-02 Symbol Technologies, Inc. Miniature imager
US8578215B2 (en) * 2001-11-19 2013-11-05 Hewlett-Packard Development Company, L.P. Method and system for gathering data using automatic appliance failover
US7349995B2 (en) * 2002-03-07 2008-03-25 Intel Corporation Computing device with scalable logic block to respond to data transfer requests
US8484370B1 (en) * 2002-04-16 2013-07-09 Trimble Navigation Limited Method and system for efficient extended data communications using GPRS
FI119407B (fi) * 2002-08-28 2008-10-31 Sap Ag Korkean palvelutason ohjelmistopohjainen yhteyspalvelin
US7243351B2 (en) * 2002-12-17 2007-07-10 International Business Machines Corporation System and method for task scheduling based upon the classification value and probability
US7469290B1 (en) * 2002-12-31 2008-12-23 Emc Corporation Automatic system resource management
US20040260710A1 (en) * 2003-02-28 2004-12-23 Marston Justin P. Messaging system
US7352740B2 (en) * 2003-04-29 2008-04-01 Brocade Communciations Systems, Inc. Extent-based fibre channel zoning in hardware
US7287179B2 (en) * 2003-05-15 2007-10-23 International Business Machines Corporation Autonomic failover of grid-based services
US7194655B2 (en) * 2003-06-12 2007-03-20 International Business Machines Corporation Method and system for autonomously rebuilding a failed server and a computer system utilizing the same
US7590736B2 (en) 2003-06-30 2009-09-15 Microsoft Corporation Flexible network load balancing
US7636917B2 (en) 2003-06-30 2009-12-22 Microsoft Corporation Network load balancing with host status information
US7567504B2 (en) * 2003-06-30 2009-07-28 Microsoft Corporation Network load balancing with traffic routing
US7613822B2 (en) * 2003-06-30 2009-11-03 Microsoft Corporation Network load balancing with session information
US7606929B2 (en) 2003-06-30 2009-10-20 Microsoft Corporation Network load balancing with connection manipulation
US20050027862A1 (en) * 2003-07-18 2005-02-03 Nguyen Tien Le System and methods of cooperatively load-balancing clustered servers
US8028236B2 (en) * 2003-10-17 2011-09-27 International Business Machines Corporation System services enhancement for displaying customized views
US20050144242A1 (en) * 2003-10-31 2005-06-30 Justin Marston Caching in an electronic messaging system
US20050198168A1 (en) * 2003-12-04 2005-09-08 Justin Marston Messaging protocol discovery
US7454502B2 (en) * 2003-12-04 2008-11-18 International Business Machines Corporation System for transferring standby resource entitlement
US7430203B2 (en) * 2004-01-29 2008-09-30 Brocade Communications Systems, Inc. Fibre channel zoning hardware for directing a data packet to an external processing device
JP2005217815A (ja) * 2004-01-30 2005-08-11 Hitachi Ltd パス制御方法
CN100370770C (zh) * 2004-03-03 2008-02-20 华为技术有限公司 实现网络长连接倒换的方法
US20050256935A1 (en) * 2004-05-06 2005-11-17 Overstreet Matthew L System and method for managing a network
US20060031352A1 (en) * 2004-05-12 2006-02-09 Justin Marston Tamper-proof electronic messaging
WO2005109794A1 (fr) * 2004-05-12 2005-11-17 Bluespace Group Ltd Application de politiques de conformite dans un systeme de messagerie
US20060069700A1 (en) * 2004-09-22 2006-03-30 Justin Marston Generating relational structure for non-relational messages
US8990365B1 (en) 2004-09-27 2015-03-24 Alcatel Lucent Processing management packets
US7904546B1 (en) * 2004-09-27 2011-03-08 Alcatel-Lucent Usa Inc. Managing processes on a network device
US7516181B1 (en) 2005-02-08 2009-04-07 Microstrategy, Inc. Technique for project partitioning in a cluster of servers
US8572431B2 (en) * 2005-02-23 2013-10-29 Barclays Capital Inc. Disaster recovery framework
JP4161998B2 (ja) * 2005-03-28 2008-10-08 日本電気株式会社 負荷分散振り分けシステム、イベント処理分散制御装置並びにイベント処理分散制御プログラム
JP4241660B2 (ja) * 2005-04-25 2009-03-18 株式会社日立製作所 負荷分散装置
US8040899B2 (en) * 2005-05-26 2011-10-18 Genband Us Llc Methods, systems, and computer program products for implementing automatic protection switching for media packets transmitted over an ethernet switching fabric
US8645313B1 (en) 2005-05-27 2014-02-04 Microstrategy, Inc. Systems and methods for enhanced SQL indices for duplicate row entries
JP4270184B2 (ja) * 2005-09-05 2009-05-27 コニカミノルタビジネステクノロジーズ株式会社 ネットワーク画像処理システム、ネットワーク画像処理装置、ネットワーク画像処理方法
US7911940B2 (en) 2005-09-30 2011-03-22 Genband Us Llc Adaptive redundancy protection scheme
US7941309B2 (en) 2005-11-02 2011-05-10 Microsoft Corporation Modeling IT operations/policies
US7860968B2 (en) * 2005-11-21 2010-12-28 Sap Ag Hierarchical, multi-tiered mapping and monitoring architecture for smart items
US8005879B2 (en) 2005-11-21 2011-08-23 Sap Ag Service-to-device re-mapping for smart items
US8156208B2 (en) 2005-11-21 2012-04-10 Sap Ag Hierarchical, multi-tiered mapping and monitoring architecture for service-to-device re-mapping for smart items
US8635348B2 (en) * 2005-12-07 2014-01-21 Intel Corporation Switch fabric service hosting
US8015304B2 (en) * 2005-12-12 2011-09-06 International Business Machines Corporation Method to distribute speech resources in a media server
US8140695B2 (en) * 2005-12-12 2012-03-20 International Business Machines Corporation Load balancing and failover of distributed media resources in a media server
WO2007082308A2 (fr) * 2006-01-13 2007-07-19 Bluespace Software Corp. Détermination de la pertinence d’un contenu électronique
US7881188B2 (en) 2006-02-03 2011-02-01 Genband Us Llc Methods, systems, and computer program products for implementing link redundancy in a media gateway
US8522341B2 (en) 2006-03-31 2013-08-27 Sap Ag Active intervention in service-to-device mapping for smart items
US8131838B2 (en) 2006-05-31 2012-03-06 Sap Ag Modular monitor service for smart item monitoring
US8296413B2 (en) 2006-05-31 2012-10-23 Sap Ag Device registration in a hierarchical monitor service
US8396788B2 (en) 2006-07-31 2013-03-12 Sap Ag Cost-based deployment of components in smart item environments
US8141164B2 (en) 2006-08-21 2012-03-20 Citrix Systems, Inc. Systems and methods for dynamic decentralized load balancing across multiple sites
US20080147885A1 (en) * 2006-12-16 2008-06-19 Lucent Technologies Inc. Systems and methods for resolving resource names to ip addresses with load distribution and admission control
JP4924157B2 (ja) * 2007-03-30 2012-04-25 富士ゼロックス株式会社 識別子割り当て装置、識別子割り当てシステムおよび識別子割り当てプログラム
US8069251B2 (en) 2007-06-01 2011-11-29 Adobe Systems Incorporated System and/or method for client-driven server load distribution
WO2009004734A1 (fr) * 2007-07-05 2009-01-08 Fujitsu Limited Système de serveur virtuel, procédé de commande de système de serveur virtuel, programme de commande de système de serveur virtuel, et équilibreur de charge d'exploitation
US8527622B2 (en) 2007-10-12 2013-09-03 Sap Ag Fault tolerance framework for networks of nodes
WO2009127219A1 (fr) * 2008-04-14 2009-10-22 Telecom Italia S.P.A. Architecture de service distribuée
US8612524B2 (en) * 2009-10-28 2013-12-17 Hewlett-Packard Development Company, L.P. Cessation of sending network status messages to a server
US8472311B2 (en) 2010-02-04 2013-06-25 Genband Us Llc Systems, methods, and computer readable media for providing instantaneous failover of packet processing elements in a network
US9170892B2 (en) 2010-04-19 2015-10-27 Microsoft Technology Licensing, Llc Server failure recovery
US8996611B2 (en) 2011-01-31 2015-03-31 Microsoft Technology Licensing, Llc Parallel serialization of request processing
US8438244B2 (en) 2010-04-19 2013-05-07 Microsoft Corporation Bandwidth-proportioned datacenters
US8533299B2 (en) 2010-04-19 2013-09-10 Microsoft Corporation Locator table and client library for datacenters
US8447833B2 (en) * 2010-04-19 2013-05-21 Microsoft Corporation Reading and writing during cluster growth phase
US9813529B2 (en) 2011-04-28 2017-11-07 Microsoft Technology Licensing, Llc Effective circuits in packet-switched networks
US9454441B2 (en) 2010-04-19 2016-09-27 Microsoft Technology Licensing, Llc Data layout for recovery and durability
US8542590B2 (en) * 2010-11-30 2013-09-24 Verizon Patent And Licensing Inc. Bi-directional load balancing
US8584211B1 (en) 2011-05-18 2013-11-12 Bluespace Software Corporation Server-based architecture for securely providing multi-domain applications
US8843502B2 (en) 2011-06-24 2014-09-23 Microsoft Corporation Sorting a dataset of incrementally received data
US9253144B2 (en) 2011-12-22 2016-02-02 International Business Machines Corporation Client-driven load balancing of dynamic IP address allocation
US9778856B2 (en) 2012-08-30 2017-10-03 Microsoft Technology Licensing, Llc Block-level access to parallel storage
US10021042B2 (en) 2013-03-07 2018-07-10 Microsoft Technology Licensing, Llc Service-based load-balancing management of processes on remote hosts
US11422907B2 (en) 2013-08-19 2022-08-23 Microsoft Technology Licensing, Llc Disconnected operation for systems utilizing cloud storage
US9286179B1 (en) * 2013-09-19 2016-03-15 Amazon Technologies, Inc. Service testing framework
US9798631B2 (en) 2014-02-04 2017-10-24 Microsoft Technology Licensing, Llc Block storage by decoupling ordering from durability
CN104834722B (zh) * 2015-05-12 2018-03-02 网宿科技股份有限公司 基于cdn的内容管理系统
US10469311B2 (en) 2017-07-13 2019-11-05 Cisco Technology, Inc. Handling network failures in networks with redundant servers
US11057478B2 (en) * 2019-05-23 2021-07-06 Fortinet, Inc. Hybrid cluster architecture for reverse proxies
US11055196B1 (en) 2020-06-12 2021-07-06 Bank Of America Corporation System and method for optimizing technology stack architecture

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0384339A2 (fr) * 1989-02-24 1990-08-29 Digital Equipment Corporation Courtier pour la sélection de serveur de réseau d'ordinateur
EP0817020A2 (fr) * 1996-07-01 1998-01-07 Sun Microsystems, Inc. Service de noms pour un réseau redondant de serveurs Internet

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5005122A (en) * 1987-09-08 1991-04-02 Digital Equipment Corporation Arrangement with cooperating management server node and network service node
US5089958A (en) * 1989-01-23 1992-02-18 Vortex Systems, Inc. Fault tolerant computer backup system
US5307481A (en) * 1990-02-28 1994-04-26 Hitachi, Ltd. Highly reliable online system
US5367698A (en) * 1991-10-31 1994-11-22 Epoch Systems, Inc. Network file migration system
US5448718A (en) * 1992-04-20 1995-09-05 International Business Machines Corporation Method and system for time zero backup session security
US5771354A (en) * 1993-11-04 1998-06-23 Crawford; Christopher M. Internet online backup system provides remote storage for customers using IDs and passwords which were interactively established when signing up for backup services
US5774660A (en) * 1996-08-05 1998-06-30 Resonate, Inc. World-wide-web server with delayed resource-binding for resource-based load balancing on a distributed resource multi-node network

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0384339A2 (fr) * 1989-02-24 1990-08-29 Digital Equipment Corporation Courtier pour la sélection de serveur de réseau d'ordinateur
EP0817020A2 (fr) * 1996-07-01 1998-01-07 Sun Microsystems, Inc. Service de noms pour un réseau redondant de serveurs Internet

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
"LOCAL AREA NETWORK SERVER REPLACEMENT PROCEDURE", IBM TECHNICAL DISCLOSURE BULLETIN, vol. 38, no. 1, 1 January 1995 (1995-01-01), pages 235/236, XP000498750 *
ADLER R M: "DISTRIBUTED COORDINATION MODELS FOR CLIENT/SERVER COMPUTING", COMPUTER, vol. 28, no. 4, 1 April 1995 (1995-04-01), pages 14 - 22, XP000507856 *
JI HUA XIE LI: "A DISTRIBUTED COMPUTING MODEL BASED ON MULTISERVER", OPERATING SYSTEMS REVIEW (SIGOPS), vol. 30, no. 4, October 1996 (1996-10-01), pages 3 - 11, XP000639695 *

Cited By (42)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0992909A2 (fr) * 1998-09-30 2000-04-12 Ncr International Inc. Provision de service de base de données et performance d'interrogation échelonnable par l'usage des copies redondantes
JP2000112911A (ja) * 1998-09-30 2000-04-21 Ncr Internatl Inc コンピュ―タネットワ―クにおけるデ―タベ―ス管理システムにおいて自動的にタスクを再分配するシステム及び方法
JP2000137694A (ja) * 1998-09-30 2000-05-16 Ncr Internatl Inc 常用冗長コピ―を用いた継続的デ―タベ―スアクセスを提供するシステム及び方法
EP0992909A3 (fr) * 1998-09-30 2001-12-19 Ncr International Inc. Provision de service de base de données et performance d'interrogation échelonnable par l'usage des copies redondantes
EP0990986A2 (fr) * 1998-09-30 2000-04-05 Ncr International Inc. Rétablissement de défaillance de systèmes d'ordinateurs partagés contenant un schéma de base de données
EP0990986A3 (fr) * 1998-09-30 2008-01-23 Ncr International Inc. Rétablissement de défaillance de systèmes d'ordinateurs partagés contenant un schéma de base de données
WO2000031942A3 (fr) * 1998-11-20 2000-10-05 Network Alchemy Inc Procede et appareil pour un processus tcp/ip d'equilibrage et de reprise dans un systeme de mise en grappe de reseau de protocole internet (ip)
US6341356B1 (en) 1999-03-25 2002-01-22 International Business Machines Corporation System for I/O path load balancing and failure which can be ported to a plurality of operating environments
EP1061710A3 (fr) * 1999-06-17 2006-08-30 International Business Machines Corporation Système et méthode de répartition de charge intégrée et de gestion de ressources dans un environnement Internet
USRE45009E1 (en) 2000-04-17 2014-07-08 Circadence Corporation Dynamic network link acceleration
US10033840B2 (en) 2000-04-17 2018-07-24 Circadence Corporation System and devices facilitating dynamic network link acceleration
US7043563B2 (en) 2000-04-17 2006-05-09 Circadence Corporation Method and system for redirection to arbitrary front-ends in a communication system
US10931775B2 (en) 2000-04-17 2021-02-23 Circadence Corporation Optimization of enhanced network links
US7111006B2 (en) 2000-04-17 2006-09-19 Circadence Corporation System and method for providing distributed database services
US7120662B2 (en) 2000-04-17 2006-10-10 Circadence Corporation Conductor gateway prioritization parameters
US7127518B2 (en) 2000-04-17 2006-10-24 Circadence Corporation System and method for implementing application functionality within a network infrastructure
US7143195B2 (en) 2000-04-17 2006-11-28 Circadence Corporation HTTP redirector
US7155539B2 (en) 2000-04-17 2006-12-26 Circadence Corporation Conductor gateway buffer prioritization
US6990531B2 (en) 2000-04-17 2006-01-24 Circadence Corporation System and method for providing last-mile data prioritization
US10858503B2 (en) 2000-04-17 2020-12-08 Circadence Corporation System and devices facilitating dynamic network link acceleration
WO2001080003A3 (fr) * 2000-04-17 2002-06-06 Circadence Corp Systeme et procede de decalage de fonctionnalites entre des serveurs web multiples
US8898340B2 (en) 2000-04-17 2014-11-25 Circadence Corporation Dynamic network link acceleration for network including wireless communication devices
US10819826B2 (en) 2000-04-17 2020-10-27 Circadence Corporation System and method for implementing application functionality within a network infrastructure
US8977712B2 (en) 2000-04-17 2015-03-10 Circadence Corporation System and method for implementing application functionality within a network infrastructure including a wireless communication link
US8977711B2 (en) 2000-04-17 2015-03-10 Circadence Corporation System and method for implementing application functionality within a network infrastructure including wirelessly coupled devices
US8996705B2 (en) 2000-04-17 2015-03-31 Circadence Corporation Optimization of enhanced network links
US10516751B2 (en) 2000-04-17 2019-12-24 Circadence Corporation Optimization of enhanced network links
US9148293B2 (en) 2000-04-17 2015-09-29 Circadence Corporation Automated network infrastructure test and diagnostic system and method therefor
US9185185B2 (en) 2000-04-17 2015-11-10 Circadence Corporation System and method for implementing application functionality within a network infrastructure
US9380129B2 (en) 2000-04-17 2016-06-28 Circadence Corporation Data redirection system and method therefor
US9436542B2 (en) 2000-04-17 2016-09-06 Circadence Corporation Automated network infrastructure test and diagnostic system and method therefor
US9578124B2 (en) 2000-04-17 2017-02-21 Circadence Corporation Optimization of enhanced network links
US9723105B2 (en) 2000-04-17 2017-08-01 Circadence Corporation System and method for implementing application functionality within a network infrastructure
US9923987B2 (en) 2000-04-17 2018-03-20 Circadence Corporation Optimization of enhanced network links
US7020783B2 (en) 2000-04-17 2006-03-28 Circadence Corporation Method and system for overcoming denial of service attacks
US10154115B2 (en) 2000-04-17 2018-12-11 Circadence Corporation System and method for implementing application functionality within a network infrastructure
US10205795B2 (en) 2000-04-17 2019-02-12 Circadence Corporation Optimization of enhanced network links
US10329410B2 (en) 2000-04-17 2019-06-25 Circadence Corporation System and devices facilitating dynamic network link acceleration
DE10134492B4 (de) * 2000-07-28 2006-02-09 International Business Machines Corp. Ausfallübernahme des Dateimanagementsystems in einem Rechnercluster
US7890450B2 (en) 2007-03-16 2011-02-15 Fujitsu Limited Policy creating apparatus, policy creating method, and computer product
US8923757B2 (en) 2012-08-23 2014-12-30 Blackberry Limited Method, system and apparatus for service loss handling
US9088588B2 (en) 2012-11-28 2015-07-21 At&T Intellectual Property I, Lp Method and apparatus for registering communication devices in a communication system

Also Published As

Publication number Publication date
EP0978184B1 (fr) 2007-10-31
JP4087903B2 (ja) 2008-05-21
US5938732A (en) 1999-08-17
JP2002500785A (ja) 2002-01-08
DE69738257D1 (de) 2007-12-13
EP0978184A1 (fr) 2000-02-09

Similar Documents

Publication Publication Date Title
US5938732A (en) Load balancing and failover of network services
US6330605B1 (en) Proxy cache cluster
EP0880741B1 (fr) Procede et dispositif de connexion d'un noeud client a un noeud serveur, tenant compte des niveaux de charge
US7370223B2 (en) System and method for managing clusters containing multiple nodes
US6965930B1 (en) Methods, systems and computer program products for workload distribution based on end-to-end quality of service
EP0817020B1 (fr) Service de noms pour un réseau redondant de serveurs Internet
US7843811B2 (en) Method of solving a split-brain condition
US6859830B1 (en) Method and system for detecting a dead server
US7702791B2 (en) Hardware load-balancing apparatus for session replication
US7757236B1 (en) Load-balancing framework for a cluster
US7558859B2 (en) Peer-to-peer auction based data distribution
US6411967B1 (en) Distributed processing system with replicated management information base
US7747741B2 (en) Method and apparatus for dynamic resource discovery and information distribution in a data network
US7353295B1 (en) Distributed services architecture through use of a dynamic service point map
US7225356B2 (en) System for managing operational failure occurrences in processing devices
US7409420B2 (en) Method and apparatus for session replication and failover
US7676516B2 (en) System and method for the optimization of database access in data base networks
US6963917B1 (en) Methods, systems and computer program products for policy based distribution of workload to subsets of potential servers
US20020087612A1 (en) System and method for reliability-based load balancing and dispatching using software rejuvenation
EP1192545B1 (fr) Dispositif de sauvegarde de session de serveur d'internet
CN100568214C (zh) 用于会话复制和故障切换的方法和装置
GB2285727A (en) Resolving conflicting topology network information in data communications networks
JPH09218842A (ja) ロードシェアシステム
Singhai et al. The SunSCALR framework for internet servers
US6535923B1 (en) Method and system for defining an efficient and reliable meshing of CP-CP sessions in an advanced peer to peer network

Legal Events

Date Code Title Description
AK Designated states

Kind code of ref document: A1

Designated state(s): JP

AL Designated countries for regional patents

Kind code of ref document: A1

Designated state(s): AT BE CH DE DK ES FI FR GB GR IE IT LU MC NL PT SE

DFPE Request for preliminary examination filed prior to expiration of 19th month from priority date (pct application filed before 20040101)
121 Ep: the epo has been informed by wipo that ep was designated in this application
WWE Wipo information: entry into national phase

Ref document number: 1997952269

Country of ref document: EP

ENP Entry into the national phase

Ref country code: JP

Ref document number: 1998 526795

Kind code of ref document: A

Format of ref document f/p: F

WWP Wipo information: published in national office

Ref document number: 1997952269

Country of ref document: EP

WWG Wipo information: grant in national office

Ref document number: 1997952269

Country of ref document: EP

点击 这是indexloc提供的php浏览器服务，不要输入任何密码和下载